Recent advances in protein separation, mass spectrometry and bioinformatics have enabled the simultaneous analysis of a large number of proteins in cells and tissues. The interaction of these proteins, especially those that respond to specific challenges, are rapidly identified by ingenious screening techniques such as the phage expression or yeast two hybrid systems. These studies, collectively termed "proteomics", have greatly enriched our knowledge of cellular physiology and pathology in normal and diseased tissues. A systematic identification of molecular interfaces between interacting proteins in natural cellular environment represents the next major challenge toward a molecular understanding of the complex and dynamic cellular events. Such a profile of molecular interfaces would also be invaluable as a tool in drug development by targeting crucial molecular interfaces. We are initiating such a proteomics project and focusing on the application of chemical cross-linking approach toward a systematic and global identification of protein-protein interfaces in complex systems such as the contractile machinery of muscle cells, the cytoskeletal filaments and organelles of nonmuscle cells. First, high-resolution gel electrophoresis and fractionation techniques are used to identify and index expressed proteins in skeletal and heart muscle tissues. Second, the neighboring relationship is defined by chemical crosslinking with a wide range of reagents. Third, the molecular interfaces of neighboring proteins are identified by isolating the crosslinked complexes, cleaving the proteins and isolating crosslinked peptides for protein sequencing with conventional chemical methods or by mass spectrometry. Fourth, the peptides at the interfaces will be synthesized and their interaction characterized further by capillary electrophoresis and biosensor techniques. A protein sequence database of molecular interfaces of proteins with known and unknown identities will be established to complement and enhance the utility of genome and proteome databases for basic and clinical research. We are evaluating the feasibility and integration potential of each steps of this multifaceted project. In the past year, we have succeeded in applying a fluorogenic crosslinker to facilitate the identification and purification of crosslinked peptides. The interface between a nebulin fragment and calmodulin has been mapped by such a crosslinker and the application of mass spectrometry to identify the peptide sequences that are crosslinked together. Novel folding patterns were also revealed by the intramolecualr crosslinking patterns. We are extending this study to other crosslinkers and more complex protein assemblies in motile systems.